Interactive Identification Method for Box-Jenkins Models

This paper converts a Box-Jenkins model into two identification submodels with the system model parameters and the noise model parameters, respectively. However, the information vectors in the submodels contain unmeasurable variables, which leads the conventional recursive least squares algorithm impossible to generate the parameter estimates. In order to overcome this difficulty, the interactive least squares algorithm is derived by using the auxiliary model identification idea and the hierarchical identification principle. The simulation results indicate that the proposed algorithm has less computational burden and more accurate parameter estimation compared with the auxiliary model based recursive generalized extended least squares algorithm.